# Importing the required libraries
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.graph_objects as go
from scipy import stats
# Loading Data
stocks = pd.read_csv('data/big_tech_stock_prices.csv')
companies = pd.read_csv('data/big_tech_companies.csv')Big Tech Stock Prices
Analyzing Big Tech stock prices from 2010 to 2022
Abstract
The goal of this project is to understand the temporal behavior of Big Tech stocks that are actively traded on the NYSE. This analysis will show trends in the market to understand times when the market was on a downtrend (bear market) or in an uptrend (bull market). This will provide insight into times when investments should be made or when they should be pulled back. Investment analyses such as the aformentioned are extremely valuable to large finanacial institutions that aim to benefit from growth in the market. This is extremely important for those who place their hard earned dollar with these institutions as a means to save for retirement.
The analysis will consist of answering two key questions in a sequential manner: exploratory data analysis (EDA), data wrangling, and data visualization. This process is key to gaining insight from the data and allows for a clean platform that can used to train machine learning models.
Q1: How do stock prices change over time, looking at the basic information like open, close, high, low etc.**
EDA
# Step 1: Exploratory Data Analysis (EDA)
df = stocks.copy()
# Convert the date column to datetime format
df['date'] = pd.to_datetime(df['date'])
# Summary Statistics
company_stats = df.groupby('stock_symbol').agg({'open': ['mean', 'median', 'min', 'max', 'std'], 'volume': ['mean', 'median', 'min', 'max', 'std']})
print("Summary Statistics:")
print(company_stats)Summary Statistics:
open \
mean median min max std
stock_symbol
AAPL 51.274171 29.745001 6.870357 182.630005 47.316509
ADBE 186.023938 97.589996 22.969999 696.280029 173.562115
AMZN 58.937204 36.325001 5.296500 187.199997 54.138607
CRM 103.419948 76.290001 15.522500 310.000000 71.287894
CSCO 33.493880 29.500000 13.930000 64.040001 12.643887
GOOGL 49.149625 38.521000 10.968719 151.250000 35.809816
IBM 148.427471 143.173996 90.439774 205.908218 23.998357
INTC 36.481406 34.270000 17.879999 68.199997 12.912534
META 147.874275 141.620002 18.080000 381.679993 86.734656
MSFT 100.048490 55.660000 23.090000 344.619995 88.228008
NFLX 188.242206 110.010002 6.960000 692.349976 178.967463
NVDA 50.560335 11.902500 2.180000 335.170013 69.538684
ORCL 46.242177 41.750000 21.459999 104.290001 16.864193
TSLA 58.859467 16.229000 1.076000 411.470001 95.677282
volume
mean median min max std
stock_symbol
AAPL 2.563255e+08 166674000.0 35195900 1880998000 2.225768e+08
ADBE 3.814337e+06 2948500.0 589200 108752400 3.598144e+06
AMZN 8.833999e+07 74592000.0 17626000 848422000 5.309249e+07
CRM 6.910973e+06 5548800.0 1084700 64562800 5.048860e+06
CSCO 3.269656e+07 25482400.0 5720500 560040200 2.570963e+07
GOOGL 6.018647e+07 41234000.0 9312000 592399008 4.957963e+07
IBM 5.036545e+06 4345189.0 1247878 39814421 2.772073e+06
INTC 3.607170e+07 29874600.0 5893800 199002600 2.123178e+07
META 3.117815e+07 23239000.0 5913100 573576400 2.713267e+07
MSFT 3.801647e+07 32280800.0 7425600 319317900 2.147328e+07
NFLX 1.841485e+07 11961800.0 1144000 315541800 2.054316e+07
NVDA 5.080613e+07 43395600.0 4564400 369292800 3.210953e+07
ORCL 1.801856e+07 14699800.0 2754900 183503900 1.251053e+07
TSLA 9.351647e+07 75914250.0 1777500 914082000 8.164780e+07 Data Wrangling
# Step 2: Data Wrangling
# Handle missing values(No missing values)Data Visualization
# Initialize an empty list to store candlestick traces for all companies
candlestick_traces = []
# Loop through each company and create a candlestick trace
for company in stocks['stock_symbol'].unique():
df_company = stocks[stocks['stock_symbol'] == company]
candlestick_trace = go.Candlestick(x=df_company['date'],
open=df_company['open'],
high=df_company['high'],
low=df_company['low'],
close=df_company['close'],
name=f"{company} Candlestick")
candlestick_traces.append(candlestick_trace)
# Create the figure
fig = go.Figure(data=candlestick_traces)
# Customize the layout
fig.update_layout(title='Candlestick Chart for 14 Companies',
xaxis_title='Date',
yaxis_title='Price',
xaxis_rangeslider_visible=False,
height=600,
width=1000)
# Show the figure
fig.show()Closing Prices Over Time for Different Companies
stocks = companies['stock_symbol'].values.tolist()
# Creating a DataFrame to hold all the stock data
# For simplicity, let's generate some random closing prices
all_stock_data = df
# Create a Plotly figure
fig = go.Figure()
# Add a line for each stock
for stock in stocks:
stock_data = df[df['stock_symbol'] == stock]
fig.add_trace(go.Scatter(x=stock_data['date'], y=stock_data['close'], mode='lines', name=stock))
# Customize the layout
fig.update_layout(
title='Stock Close Prices Over Time',
xaxis_title='Date',
yaxis_title='Close Price',
height=400
)
# Show the figure
fig.show()Volume Analysis using Bar Plot
# Calculate average volume traded for each stock symbol
avg_volume = df.groupby('stock_symbol')['volume'].mean().sort_values()
# Create a bar trace
bar_trace = go.Bar(
x=avg_volume.index, # Stock symbols
y=avg_volume.values, # Average volume
marker_color='skyblue'
)
# Create a layout
layout = go.Layout(
title='Average Volume Traded for Each Stock Symbol',
xaxis=dict(title='Stock Symbol'),
yaxis=dict(title='Average Volume (Shares Traded)'),
xaxis_tickangle=-45,
margin=dict(l=40, r=40, t=60, b=20)
)
# Create a figure
fig = go.Figure(data=[bar_trace], layout=layout)
# Show the figure
fig.show()Q2 Backwards verification: if we invested x amount of dollars in 2010, how much would it be worth in 2022, when would be a good/bad time to pull investment out of the market.
EDA
# Load the CSV files
stocks = pd.read_csv('data/big_tech_stock_prices.csv')
companies = pd.read_csv('data/big_tech_companies.csv')
# Copy the DataFrames to avoid modifying the original data
df_stocks = stocks.copy()
df_companies = companies.copy()
# Now you can proceed with your groupby aggregation
investment_stats = df_stocks.groupby('stock_symbol').agg({
'open': ['mean', 'median', 'min', 'max', 'std'],
'high': ['mean', 'median', 'min', 'max', 'std'],
'low': ['mean', 'median', 'min', 'max', 'std'],
'close': ['mean', 'median', 'min', 'max', 'std'],
'adj_close': ['mean', 'median', 'min', 'max', 'std'],
'volume': ['mean', 'median', 'min', 'max', 'std']
})
print("Summary Statistics:")
print(investment_stats)Summary Statistics:
open \
mean median min max std
stock_symbol
AAPL 51.274171 29.745001 6.870357 182.630005 47.316509
ADBE 186.023938 97.589996 22.969999 696.280029 173.562115
AMZN 58.937204 36.325001 5.296500 187.199997 54.138607
CRM 103.419948 76.290001 15.522500 310.000000 71.287894
CSCO 33.493880 29.500000 13.930000 64.040001 12.643887
GOOGL 49.149625 38.521000 10.968719 151.250000 35.809816
IBM 148.427471 143.173996 90.439774 205.908218 23.998357
INTC 36.481406 34.270000 17.879999 68.199997 12.912534
META 147.874275 141.620002 18.080000 381.679993 86.734656
MSFT 100.048490 55.660000 23.090000 344.619995 88.228008
NFLX 188.242206 110.010002 6.960000 692.349976 178.967463
NVDA 50.560335 11.902500 2.180000 335.170013 69.538684
ORCL 46.242177 41.750000 21.459999 104.290001 16.864193
TSLA 58.859467 16.229000 1.076000 411.470001 95.677282
high ... \
mean median min max std ...
stock_symbol ...
AAPL 51.845876 29.980000 7.000000 182.940002 47.926721 ...
ADBE 188.208985 98.239998 23.360001 699.539978 175.674664 ...
AMZN 59.610805 36.500000 5.564500 188.654007 54.785319 ...
CRM 104.720219 77.160004 15.625000 311.750000 72.149412 ...
CSCO 33.804014 29.770000 14.120000 64.290001 12.760317 ...
GOOGL 49.638485 38.930000 11.068068 151.546494 36.215672 ...
IBM 149.571376 144.160004 93.441681 206.405350 24.013109 ...
INTC 36.875197 34.580002 17.920000 69.290001 13.090012 ...
META 149.757980 143.415000 18.270000 384.329987 87.818979 ...
MSFT 101.039621 56.000000 23.320000 349.670013 89.181105 ...
NFLX 191.178020 111.900002 7.178571 700.989990 181.488846 ...
NVDA 51.504697 11.982500 2.262500 346.470001 70.979752 ...
ORCL 46.696918 42.000000 21.680000 106.339996 17.059092 ...
TSLA 60.174863 16.491000 1.108667 414.496674 97.873400 ...
adj_close \
mean median min max std
stock_symbol
AAPL 49.445122 27.385101 5.846675 180.959732 47.810585
ADBE 186.022299 97.720001 22.690001 688.369995 173.466083
AMZN 58.905287 36.382500 5.430500 186.570496 54.085034
CRM 103.400510 76.260002 15.520000 309.959991 71.213839
CSCO 28.624463 23.570276 9.743538 61.521923 13.590111
GOOGL 49.148954 38.538502 10.912663 149.838501 35.803048
IBM 113.148579 113.695961 75.138626 150.570007 14.336266
INTC 31.310983 28.492294 12.135988 64.383247 13.745841
META 147.913244 142.065002 17.730000 382.179993 86.763257
MSFT 95.285446 50.052330 17.769510 339.924835 89.253596
NFLX 188.252178 110.099998 7.018571 691.690002 178.877130
NVDA 50.282939 11.685297 2.037410 333.407379 69.495321
ORCL 42.572566 37.007454 17.991089 101.501656 17.867223
TSLA 58.805222 16.222334 1.053333 409.970001 95.544413
volume
mean median min max std
stock_symbol
AAPL 2.563255e+08 166674000.0 35195900 1880998000 2.225768e+08
ADBE 3.814337e+06 2948500.0 589200 108752400 3.598144e+06
AMZN 8.833999e+07 74592000.0 17626000 848422000 5.309249e+07
CRM 6.910973e+06 5548800.0 1084700 64562800 5.048860e+06
CSCO 3.269656e+07 25482400.0 5720500 560040200 2.570963e+07
GOOGL 6.018647e+07 41234000.0 9312000 592399008 4.957963e+07
IBM 5.036545e+06 4345189.0 1247878 39814421 2.772073e+06
INTC 3.607170e+07 29874600.0 5893800 199002600 2.123178e+07
META 3.117815e+07 23239000.0 5913100 573576400 2.713267e+07
MSFT 3.801647e+07 32280800.0 7425600 319317900 2.147328e+07
NFLX 1.841485e+07 11961800.0 1144000 315541800 2.054316e+07
NVDA 5.080613e+07 43395600.0 4564400 369292800 3.210953e+07
ORCL 1.801856e+07 14699800.0 2754900 183503900 1.251053e+07
TSLA 9.351647e+07 75914250.0 1777500 914082000 8.164780e+07
[14 rows x 30 columns]Data Wrangling
# Handle missing values(No missing values)
print("Null Values from Stocks dataset", stocks.isnull().sum())
print("Null Values from companies dataset", companies.isnull().sum())
# Outlier Detection and Treatment
# Calculate the Z-score for the 'adj_close' column
df_stocks['z_score'] = stats.zscore(df_stocks['adj_close'])
# Remove outliers beyond 3 standard deviations
df_stocks = df_stocks[df_stocks['z_score'].abs() <= 3]Null Values from Stocks dataset stock_symbol 0
date 0
open 0
high 0
low 0
close 0
adj_close 0
volume 0
dtype: int64
Null Values from companies dataset stock_symbol 0
company 0
dtype: int64Data Visualization
stocks = companies['stock_symbol'].values.tolist()
# Creating a DataFrame to hold all the stock data
all_stock_data = df_stocks
# Create a Plotly figure
fig = go.Figure()
# Add a line for each stock
for stock in stocks:
stock_data = df_stocks[df_stocks['stock_symbol'] == stock]
fig.add_trace(go.Scatter(x=stock_data['date'], y=stock_data['close'], mode='lines', name=stock))
# Customize the layout
fig.update_layout(
title='Stock Close Prices Over Time',
xaxis_title='Date',
yaxis_title='Close Price',
height=800
)
# Show the figure
fig.show()start_date = "2010-01-01"
filtered_data = all_stock_data[all_stock_data['date'] >= start_date]
# Normalize each stock's prices to start from a base (e.g., 100)
base = 10
for stock in stocks:
initial_price = filtered_data[filtered_data['stock_symbol'] == stock].iloc[0]['close']
filtered_data.loc[filtered_data['stock_symbol'] == stock, 'Normalized Close'] = base * (filtered_data['close'] / initial_price)
# Create a Plotly figure
fig = go.Figure()
# Add a line for each stock's normalized close prices
for stock in stocks:
stock_data = filtered_data[filtered_data['stock_symbol'] == stock]
fig.add_trace(go.Scatter(x=stock_data['date'], y=stock_data['Normalized Close'], mode='lines', name=stock))
# Customize the layout
fig.update_layout(
title='Normalized Stock Close Prices Over Time',
xaxis_title='Date',
yaxis_title='Normalized Close Price',
height=600
)
# Show the figure
fig.show()To answer the question “when would be a good/bad time to pull investment out of the market?” we rely on a common trading strategy based on moving averages:
The 50-day moving average (50DMA) is often used as a short-term trend indicator. It’s more sensitive to price changes in the recent past.
The 200-day moving average (200DMA) is commonly used as a long-term trend indicator. It’s less sensitive to daily price fluctuations and more indicative of a longer-term trend.
When the 50DMA crosses above the 200DMA, it’s often referred to as a “golden cross,” which some traders interpret as a bullish signal suggesting that the price might increase, potentially a good time to buy.
Conversely, when the 50DMA crosses below the 200DMA, it’s known as a “death cross,” which is considered by some traders to be a bearish signal, indicating that the price might drop, and it could be a good time to sell.
df_AAPL = df_stocks[df_stocks['stock_symbol'] == 'AAPL']
start_date = "2013-01-01"
df_AAPL = df_AAPL[df_AAPL['date'] >= start_date]
# Calculate 200DMA and 50DMA
df_AAPL['200DMA'] = df_AAPL['close'].rolling(window=200).mean()
df_AAPL['50DMA'] = df_AAPL['close'].rolling(window=50).mean()
# Creating the candlestick chart
fig = go.Figure()
# Add Candlestick trace
fig.add_trace(go.Candlestick(x=df_AAPL['date'],
open=df_AAPL['open'], high=df_AAPL['high'],
low=df_AAPL['low'], close=df_AAPL['close'],
name='Candlestick'))
# Add 200DMA Line trace
fig.add_trace(go.Scatter(x=df_AAPL['date'], y=df_AAPL['200DMA'],
mode='lines', name='200DMA',
line=dict(color='blue', width=2)))
# Add 50DMA Line trace
fig.add_trace(go.Scatter(x=df_AAPL['date'], y=df_AAPL['50DMA'],
mode='lines', name='50DMA',
line=dict(color='red', width=2)))
# Customize the layout
fig.update_layout(title='Stock Price with 200DMA and 50DMA',
xaxis_title='Date',
yaxis_title='Price',
xaxis_rangeslider_visible=False, # Hides the range slider
height=600)
# Show the figure
fig.show()